How to Find a Heart Attack Long Before It Happens

The troponin proteins—cardiac troponin I and T—are the best biomarkers we have for determining whether there’s been damage to the heart. As many as 80% of patients who have had a heart attack will show an elevated level of cardiac troponin I in their blood within two to three hours of visiting the emergency room.

But while troponin—which is released when heart cells are damaged or die—is the most sensitive and specific marker we have for serious cardiac injury, could it also offer an early warning sign of such injury?

That has been the hope driving innovation in a slew of new assays that aim to measure troponin and a host of other telltale proteins—more on those in a bit—with extraordinary sensitivity. Ideally, if we could measure minuscule amounts of troponin in healthy individuals, we could also detect the most incremental changes in its level over time—and warn of heart failure or early ischemic disease when there is still time enough to slow or reverse it.

“Today, most of the technologies that can see proteins in blood can see disease, but they can’t see health,” says Kevin Hrusovsky, CEO and executive chairman of Quanterix, which makes an über-sensitive protein-detection machine called Simoa. “And that continuum is very important because if we can see any migration from baseline, and be able to detect things very early, we can see any trend away from health.”

Simoa works by putting 500,000 little antibody-connected beads into a sample of blood. The beads slosh around in the blood and when they bump into a specific type of protein, they connect to it like Velcro. Then Simoa tosses in a second antibody—this one tied to a fluorescent light—which gloms on to any bound protein-antibody complex from the first round. The beads are then rolled onto a Sony disc that has 212,000 little holes on it. The beads fall into the wells, which are just a tiny bit bigger, “like a bingo ball sliding into a bingo tray,” says Hrusovsky. Eventually, a laser camera scans every single well to see whether or not there’s a light on a given bead. Simoa can detect just one light on one bead in one well—which is to say a single target protein in a relative ocean of blood.

The technology is based on what’s called a “sandwich ELISA,” which isn’t new at all. ELISA (enzyme-linked immunosorbent assay) has been around for more than half a century. “We are using a very traditional ELISA test and just putting some—I call it some digital trickery—onto it,” says Hrusovsky. But what the Lexington, Mass., company has really done is to figure out how to do 500,000 tests in the same time we used to do one.

That’s hack No. 1. And while Quanterix—which was cofounded in 2007 by David Walt, cofounder of gene-sequencing biggie Illumina—is (slowly) developing a diagnostic test with French firm bioMérieux, it has made the smart move not to wait around for FDA approval, a process that can take years. In the meantime, Quanterix has gotten its $165,000 fridge-sized machine into pharma, biotech, and university laboratories across the globe—essentially letting other scientists prove its mettle and publish the results. Outside scientists have already tested the company’s assays for detecting protein biomarkers associated with concussion and traumatic brain injury, Alzheimer’s disease, various cancers, and the inflammatory process that is likely involved in several pathologies.

The strategy, you may recall, is a far cry from the black-box approach of Theranos (see Carreyrou, John, WSJ). This year, the privately held Quanterix, which is backed by Bain Capital Ventures, ARCH Venture Partners, and Flagship Ventures, among others, expects to pull in $19 million in revenue—without what Hrusovsky calls “regulatory or reimbursement risk.” That’s hack No. 2.

In the burgeoning age of digital health, it’s that second hack that entrepreneurs may want to take heed of.